/* * Copyright (C) 2019 The Android Open Source Project * * Licensed under the Apache License, Version 2.0 (the "License"); * you may not use this file except in compliance with the License. * You may obtain a copy of the License at * * http://www.apache.org/licenses/LICENSE-2.0 * * Unless required by applicable law or agreed to in writing, software * distributed under the License is distributed on an "AS IS" BASIS, * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. * See the License for the specific language governing permissions and * limitations under the License. */ #ifndef SRC_TRACE_PROCESSOR_CONTAINERS_NULLABLE_VECTOR_H_ #define SRC_TRACE_PROCESSOR_CONTAINERS_NULLABLE_VECTOR_H_ #include #include #include #include "perfetto/base/logging.h" #include "src/trace_processor/containers/row_map.h" namespace perfetto { namespace trace_processor { // A data structure which compactly stores a list of possibly nullable data. // // Internally, this class is implemented using a combination of a std::deque // with a BitVector used to store whether each index is null or not. // By default, for each null value, it only uses a single bit inside the // BitVector at a slight cost (searching the BitVector to find the index into // the std::deque) when looking up the data. template class NullableVector { private: enum class Mode { // Sparse mode is the default mode and ensures that nulls are stored using // only // a single bit (at the cost of making setting null entries to non-null // O(n)). kSparse, // Dense mode forces the reservation of space for null entries which // increases // memory usage but allows for O(1) set operations. kDense, }; public: // Creates an empty NullableVector. NullableVector() : NullableVector(Mode::kSparse) {} NullableVector(const NullableVector&) = delete; NullableVector& operator=(const NullableVector&) = delete; NullableVector(NullableVector&&) = default; NullableVector& operator=(NullableVector&&) noexcept = default; // Creates a sparse nullable vector static NullableVector Sparse() { return NullableVector(Mode::kSparse); } // Creates a dense nullable vector static NullableVector Dense() { return NullableVector(Mode::kDense); } // Returns the optional value at |idx| or std::nullopt if the value is null. std::optional Get(uint32_t idx) const { bool contains = valid_.IsSet(idx); if (mode_ == Mode::kDense) return contains ? std::make_optional(data_[idx]) : std::nullopt; return contains ? std::make_optional(data_[valid_.CountSetBits(idx)]) : std::nullopt; } // Adds the given value to the NullableVector. void Append(T val) { data_.emplace_back(val); valid_.AppendTrue(); } // Adds the given optional value to the NullableVector. void Append(std::optional val) { if (val) { Append(*val); } else { AppendNull(); } } // Sets the value at |idx| to the given |val|. void Set(uint32_t idx, T val) { if (mode_ == Mode::kDense) { valid_.Set(idx); data_[idx] = val; } else { // Generally, we will be setting a null row to non-null so optimize for // that path. uint32_t row = valid_.CountSetBits(idx); bool was_set = valid_.Set(idx); if (PERFETTO_UNLIKELY(was_set)) { data_[row] = val; } else { data_.insert(data_.begin() + static_cast(row), val); } } } // Requests the removal of unused capacity. // Matches the semantics of std::vector::shrink_to_fit. void ShrinkToFit() { data_.shrink_to_fit(); valid_.ShrinkToFit(); } // Returns the size of the NullableVector; this includes any null values. uint32_t size() const { return valid_.size(); } // Returns whether data in this NullableVector is stored densely. bool IsDense() const { return mode_ == Mode::kDense; } const std::vector& non_null_vector() const { return data_; } const BitVector& non_null_bit_vector() const { return valid_; } private: explicit NullableVector(Mode mode) : mode_(mode) {} void AppendNull() { if (mode_ == Mode::kDense) { data_.emplace_back(); } valid_.AppendFalse(); } Mode mode_ = Mode::kSparse; std::vector data_; BitVector valid_; }; } // namespace trace_processor } // namespace perfetto #endif // SRC_TRACE_PROCESSOR_CONTAINERS_NULLABLE_VECTOR_H_